Harry Seldon's blog

Fractals, Chaos, and Control Systems on Rails


2 lessons Economics should learn from Aerospace

Posted by Harry Seldon on October 30, 2009

Let’s imagine that economy is an airplane. This airplane goes through a severe zone of turbulence, transforming the cruise in an heavily uncomfortable bumpy ride. Passengers start to complain, and feel more and more insecure. One of them finally calls the stewardess and asks: “What are the pilots doing ? Can’t they move us out of this unbearable situation ?” “Well… there is actually no pilot in this plane”, the stewardess answers politely. The passenger now gets really nervous, and says “What ??? Then we should run to the cockpit and try to do something before it is too late !”. The stewardess, feeling really sorry, replies with a “Uh, I’m afraid there is no cockpit in this plane, Mister”. And the story ends there, with the krach of the “economy” airplane.

I had read this little comparison a year ago already in a French newspaper. The author was Jacques Attali, a French economist. As a pilot, I have wanted to blog about this since that time.

To be a little more accurate, an analogy would assimilate companies to airplanes and market to Air Traffic Control (ATC). The problem is the market does not control anything. Imagine air traffic without ATC. Air space would be a mess. Aircrafts would collide all the time and basically nobody would trust air traffic. Finally nobody would take the plane. The air traffic system would be quickly dead. Yet, that is what is happening with the economy. Companies crash and collide because they have secant routes and nobody to help them. Economy is a mess. Most of the time ATC helps pilots by guiding them among the traffic and by giving them slots to take off and land. But sometimes ATC prevents pilots from going too fast to their destination because they would compromise safety of other airplanes. Unfortunately companies have no “market controllers” to talk to, hence the crisis.

Economy should learn 2 lessons from aerospace system engineering:

  • Systems must be controlled.
    An airplane is highly unstable. It can fly because it is very actively piloted. Air traffic is unstable too, it is controlled by ATC. In the same way economy needs guidance, navigation and control. To say it differently economy needs regulation.

  • Systems must be robustified.
    Even once an airplane is controlled and made stable there is still a lot of work to do to ensure that it is robust to all kinds of failures, even some non-anticipated failures.
    Economy needs safety engineers. Currently there is only one safety engineer for the economy. He is Nassim Nicholas Taleb, the author of the Black Swan and Benoit Mandelbrot’s spiritual son. Else financial engineers behave more like terrorists than safety engineers. OK it is a bit harsh.

I began this post by the first point. So let’s develop now the second point.
An airliner has redundancy by design. It has 3 flight computers (and more). It has 3 fly by wire circuits. It has 3 hydraulics circuits. It has 2 wings (just kidding). It has 2 pilots. To design the airplane there is a whole team of engineers whose only role is to evaluate and improve the aircraft safety. They will verify that the airplane is as black swan proof (rare catastrophic event proof) as it can be. For instance, they will check that in the case of an engine explosion the projections won’t cut all the flight control wires. Obviously, propulsion engineer (in engine companies) will make sure an engine does not explode but, all the same, aerospace safety engineers do not take anything for granted. So they study the worst cases and they robustify the system. Moreover, they also robustify the system without any specific case of failure. For instance, the logic behind having several circuits is as simple as “something could happen”. You do not always need to know the exact failure scenario to robustify the system. Obviously, nothing (and no human pilot) being perfect, there are crashes. However, it is still safer to be in an airplane than in a car.

To come back to business, notice that safety by redundancy is the 6th point in the absolutely excellent article ”The Six Mistakes Executives Make in Risk Management” by Taleb. The mistake is “We are taught that efficiency and maximizing shareholder value don’t tolerate redundancy”. (A blogger sums up the six mistakes here).
If you are a manager, don’t leverage too much your department. Don’t think that each competency must be hold by only one person. Don’t make your best to reduce all redundancies. On the contrary identify your key activities and put redundancy on it. On the short term, reducing all redundancy gives you more profits but on the medium term it leaves you exposed to very easy failures (a sick person, an expert who leaves, etc.).

To conclude I will point out that Taleb loves to compare financial analysts to blind drivers. Here is one of his last sentence: “There were so many wise people around, but whom does Obama pick? The same people who were driving a schoolbus blindfolded, who have now been given a bigger bus.” You can find the citation in this article: Black swan now elephant in the room

Posted in , | 1 comment | Tags , , | atom



Use the following link to trackback from your own site:

  1. Social comments and analytics for this post From uberVU - social comments
    This post was mentioned on Twitter by Harry Seldon: 2 lessons Economics should learn from Aerospace: If economy was an airplane, that airplane would go across a se.. http://bit.ly/3O8vhz


Leave a response

  1. Dimitar on October 24, 2010 at 11:19 am

    Hello Harry,

    A great blog you have.

    Excuse me for commenting on this article with one year delay. I was reading about Mandelbrot and one thing led to another, so I ended up here.

    I just wanted to say that redundancy won’t do any good in economics and finance. The case in view is the housing crisis. Supposedly everyone was watching it, but regulators and market participants alike failed to see what was coming, even though, with the benefit of hindsight, it looks so obvious.

    I believe incentives play a central role. Give people a good incentive and they will deny facts to support their fantasy. When the pilots, ATCs and passengers all want the plane to fly as fast as possible without regard for physics, they are bound to have crashes.

    Human systems are a whole lot different from machines, even very complex ones and social science can’t be studied or fitted to the methods of natural science. Nevertheless we have managed to improve our government and economics though all the crashes of the past. So, if we don’t wipe ourselves out, we are bound to get better at these things.

    Anyway, thank you for the book recommendations. Maybe this year you will have a new list?


Leave a comment

Markdown enabled



Recent Comments

Recent Posts


actuators aircraft atc blog chaos chaos_theory charts control controllers controls crisis economy finance flight fractals git gnc gs guidance linux mandelbrot marketing navigation ns ofc on pilot rails ruby sas scs sensors statistics systems techcrunch thinkosphere tutorial typo ubuntu wifi